LiBRA：联邦学习中基于逻辑值指纹的拜占庭鲁棒客户端选择

arXiv cs.LG 论文
federated-learning byzantine-robustness security aggregation arxiv privacy

摘要

LiBRA 提出在联邦学习的拜占庭鲁棒客户端选择中，使用客户端模型的类条件逻辑值指纹代替参数向量，实现了 Krum、Multi-Krum 和 Bulyan 的变体，在保持鲁棒性的同时大幅减少比较次数，且无需访问客户端参数。

arXiv:2610.27357v1 Announce Type: new Abstract: Byzantine-robust aggregation shields federated learning from malicious updates, but the rules that filter out suspect updates do so by comparing clients' parameter vectors. That comparison requires access to every client's update, which secure aggregation exists to withhold, costs $O(n^2 d)$ in the parameter dimension $d$, and is defined only between clients that share an architecture. We propose LiBRA, which compares clients on what their models compute rather than on the parameters that compute them, using a fingerprint of $C^2$ values for $C$ classes, the class-conditional logit response of each client's model to pseudo-patterns optimised against it. We instantiate LiBRA as variants of Krum, Multi-Krum and Bulyan that change only what the dissimilarity score is computed on. On Fashion-MNIST and CIFAR-10 with three architectures and four backdoor and label-manipulation attacks under i.i.d. partitions, each variant holds every attack its parent holds, and on identical client updates under a backdoor attack the fingerprint ranks malicious clients within 0.006 AUC of the full parameter vector. The variants thus match the robustness of the rules they extend, while their trust decision reads no client parameter, compares up to 49,090 times fewer values, and can by construction compare clients whose architectures differ, at the cost of one pattern optimisation per client per round.
查看原文
查看缓存全文

缓存时间: 2026/10/21 06:50 

# LiBRA: Logit Fingerprints for Byzantine-Robust Client Selection in Federated Learning
Source: [https://arxiv.org/abs/2610.27357](https://arxiv.org/abs/2610.27357)
[View PDF](https://arxiv.org/pdf/2610.27357)

> Abstract:Byzantine\-robust aggregation shields federated learning from malicious updates, but the rules that filter out suspect updates do so by comparing clients' parameter vectors\. That comparison requires access to every client's update, which secure aggregation exists to withhold, costs $O\(n^2 d\)$ in the parameter dimension $d$, and is defined only between clients that share an architecture\. We propose LiBRA, which compares clients on what their models compute rather than on the parameters that compute them, using a fingerprint of $C^2$ values for $C$ classes, the class\-conditional logit response of each client's model to pseudo\-patterns optimised against it\. We instantiate LiBRA as variants of Krum, Multi\-Krum and Bulyan that change only what the dissimilarity score is computed on\. On Fashion\-MNIST and CIFAR\-10 with three architectures and four backdoor and label\-manipulation attacks under i\.i\.d\. partitions, each variant holds every attack its parent holds, and on identical client updates under a backdoor attack the fingerprint ranks malicious clients within 0\.006 AUC of the full parameter vector\. The variants thus match the robustness of the rules they extend, while their trust decision reads no client parameter, compares up to 49,090 times fewer values, and can by construction compare clients whose architectures differ, at the cost of one pattern optimisation per client per round\.

## Submission history

From: Saheed Tijani \[[view email](https://arxiv.org/show-email/df02ce2a/2610.27357)\] **\[v1\]**Mon, 19 Oct 2026 22:35:44 UTC \(368 KB\)

相似文章

良性及对抗性客户端异质性下航空发动机预测的鲁棒与个性化联邦学习

arXiv cs.LG

本文针对良性及对抗性客户端异质性下的航空发动机剩余使用寿命预测，对联邦学习进行了受控研究，评估了个性化和拜占庭鲁棒聚合方法。研究发现，共享表示个性化缩小了本地与集中式准确率之间的绝大部分差距；使用Krum的鲁棒聚合能有效缓解后门攻击；将两者结合可形成组合防御，在攻击成功率较低的同时，仅付出较小的准确率代价。

现实世界客户端流失下的鲁棒联邦学习

arXiv cs.LG

FeLiX是一个新的联邦学习编排框架，通过处理临时客户端可用性、动态数据异质性和结果延迟，优化实时交互流上的目标准确率时间。它引入了流感知可用性层级、新鲜效用选择和延迟鲁棒聚合，与最先进的基线相比，将挂钟时间减少了最多2.37倍，通信带宽减少了1.30倍。